Findings from a recent study published in American Journal of Ophthalmology estimated the cumulative incidence of strabismus in children diagnosed with retinopathy of prematurity (ROP) and found a markedly higher hazard among those who underwent treatment.
The analysis pulled from a national, multicenter cohort and tracked outcomes out to five years post-diagnosis.
Give me some background first.
Some context: ROP is one of the leading causes of preventable visual impairment in children worldwide. As neonatal intensive care has improved survival of preterm infants, the at-risk population has grown alongside it.
Case in point: U.S. national data showed ROP incidence among premature infants rose from 4.4% in 2003 to 8.1% in 2019.
Now, talk about the study.
Investigators ran a retrospective cohort study using the TriNetX U.S. Collaborative network, which pools data across more than 70 healthcare organizations.
- They identified pediatric ROP patients via ICD-9 and ICD-10 codes and split the cohort by treatment exposure.
Treatment definition: Treated ROP meant receipt of laser photocoagulation and/or intravitreal (IVT) anti-vascular endothelial growth factor (anti-VEGF) injection—untreated ROP meant no record of either intervention
- Cumulative incidence was estimated at 1, 3, and 5 years using Kaplan-Meier analysis
- A multivariable Cox proportional hazards model produced adjusted hazard ratios (aHRs) accounting for demographic variables and ocular comorbidities.
Who was included in the study?
The cohort included 27,720 children aged ≤18 years with a documented ROP diagnosis between Jan. 1, 2004 and Jan. 1, 2024. Of those, 973 were treated and 26,747 were untreated.
Within the treated group: 44.81% received laser photocoagulation alone, 46.56% received anti-VEGF injections alone, and 4.42% received combined therapy.
Comorbidities ran higher in the treated arm: Amblyopia in 19.12% of treated patients versus 14.01% of untreated, and nystagmus in 8.02% versus 4.06%.
Findings?
At 5 years, cumulative incidence of any strabismus was 34.78% in the treated group versus 22.14% in the untreated group. The gap held at every analyzed timepoint:
- 1 year: 15.97% (treated) vs. 10.57% (untreated)
- 3 years: 28.55% vs. 18.72%
- 5 years: 34.78% vs. 22.14%
For comparison: In the general U.S. pediatric population, approximately 2 to 4% of children develop strabismus by age 6, which puts the ROP cohort at substantially higher risk.
Tell me more.
In the multivariable Cox model, treatment exposure was independently associated with strabismus development (aHR 1.58; 95% CI, 1.29 to 1.95; P < 0.001).
The strongest correlates: Amblyopia carried an aHR of 2.58 (95% CI, 2.25 to 2.96; P < 0.001) and nystagmus an aHR of 1.76 (95% CI, 1.48 to 2.1; P < 0.001).
- Disorders of refraction and accommodation (aHR 1.35), optic nerve and visual pathway disorders (aHR 1.55), and lens disorders (aHR 1.61) were also independently associated with elevated hazard
- Glaucoma and history of eye and orbit injury were not associated
- Sex, race, ethnicity, and gestational age likewise did not reach independent significance.
Sensitivity check: Restricting the cohort to children aged ≤4 years at index date (to address possible delayed diagnosis documentation) preserved the finding.
Treatment exposure remained associated with strabismus (aHR 1.84; 95% CI 1.06 to 3.17; P = 0.03).
Any limitations?
This was a retrospective analysis of administrative data, which carries the usual caveats. Diagnostic and procedural coding may misclassify some patients.
Granular clinical detail was not available (such as disease severity stage, treatment intensity, and refractive status)
Other considerations:
- The TriNetX network skews toward large academic medical centers, raising the possibility of selection and information bias
- Outcomes were captured only when documented within participating health systems, which may have led to an underestimation of cumulative incidence.
Expert opinion?
The authors noted that treatment exposure should be read as a clinically recognizable risk marker, since children who require laser or anti-VEGF likely had more severe baseline retinal disease.
They argued the elevated strabismus hazard in treated patients reflects greater underlying vulnerability rather than the intervention itself.
Take home.
Children with ROP, especially those who require laser or anti-VEGF therapy, accumulate strabismus risk well beyond infancy. Early ophthalmic screening alone won't catch every case.
As such: Structured longitudinal surveillance of ocular alignment through early childhood is warranted in this population, with extra attention to children with concurrent amblyopia or nystagmus.